JPH08306330A - Focal point automatic determination method - Google Patents

Abstract

PURPOSE: To provide a focal point automatic determination method which can greatly improve the accuracy of dimensional measurement and the like by an electron microscope, dispenses with a skilled operator and excels in operability. CONSTITUTION: This method comprises the following step. The scanning of a beam is performed for every excitation condition of the objective lens of an electron microscope (Steps S1, S2, S3), and the obtained secondary electron signal data is addingly averaged and movingly averaged to eliminate noise. Next, linear integralation is performed for every scanning data, thereby forming profile data expressing a linear integralation value as the function of the excitation condition. Next, the excitation condition corresponding to the peak of the profile data is sought to carry out automatic focusing. The excitation condition giving the value of a fixed rate with respect to the peak of the profile data is compared with the excitation condition giving a peak value, and the distortion of the profile data is determined from the comparison result, and when the distortion is large, the automatic focusing is determined unfocusing.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic focus determination method and, more particularly, to a sequence for determining whether or not the automatic focus adjustment of an electron microscope is successful.

[0002]

2. Description of the Related Art Conventionally, automatic focusing in an electron microscope has been performed as follows. 1) The electron beam scanning is fixed to circular scanning or line scanning and the excitation of the objective lens is changed by a constant amount.
The secondary electron signal is detected and stored. 2) Next, the integrated value of the signal change amount between the adjacent pixels or the absolute value of the difference in the signal amount between the adjacent pixels from the stored secondary electron signal. 3) Finally, the excitation condition of the objective lens that maximizes the obtained value is obtained and set as the objective lens.

The automatic focusing in the electron microscope is carried out based on the above sequence, but the operator judges visually whether the object is in focus or not.

FIG. 3 shows a pattern formed on an object to be observed, for example, a semiconductor chip, after the operator regards this as an in-focus point after the automatic focusing as described above is performed and continues to be in that state. It is the result of performing the dimension measurement of. By the way, this dimensional measurement is performed on the same pattern.

As is apparent from FIG. 3, there are cases where the measured variation in the pattern dimension exceeds ± 0.01 μm. As the measurement conditions, the dimension measurement location is the same, and the lens barrel parameters other than the automatic focusing are also the same, so this measurement value variation is because the automatic focusing state is not optimal.

For accurate measurement, it is necessary to determine whether or not the in-focus state is satisfied after performing the automatic focusing. However, as described above, this determination is conventionally made. It was done visually by the operator.

[0007]

As described above, the conventional automatic focus determination method has the following problems because it is necessary to rely on the operator's visual observation to determine whether or not to focus.

In measuring the size of a submicron pattern on a semiconductor chip, it is required that the variation caused by the device is 1/100 of the pattern design size, that is, 0.005 μm or less.

However, the dimensional measurement variation is 0.005.
A fairly skilled operator is required to visually determine the success or failure of automatic focusing so as to keep it below micron m. However, the method of leaving the determination of automatic focusing to an operator is not practical for application in an actual semiconductor manufacturing line.

In other words, there is a limit to the method of entrusting the operator's decision to determine whether the focus is acceptable or not. Therefore, if automatic focus determination is introduced, and if the focus state is determined to be a non-defective state, automatic focus adjustment is performed again. The development of such a method has been a major issue.

The object of the present invention is to solve the above-mentioned problems of the prior art and to change the excitation of the objective lens while
The next electron signal is detected, and the amount of distortion of the profile data is calculated in response to the automatic focusing performed based on the profile data of the line integral value. Based on this amount of distortion, automatic focusing is in focus or not. By determining whether it is the in-focus point, the accuracy of automatic focusing is improved, the reliability of dimension measurement with an electron microscope is greatly improved, and a skilled operator is not required. To provide.

[0012]

In order to achieve the above object, the present invention performs beam scanning for each excitation condition of an objective lens of an electron microscope and linearly integrates the obtained secondary electron signal waveform. A first step of forming profile data showing the value of the line integral as a function of the excitation condition;
A second step of obtaining a first excitation condition corresponding to the peak of the profile data, and an excitation condition corresponding to a value of a constant ratio with respect to the peak value of the profile data, a strong excitation side and a weak excitation side. The third step for obtaining the second excitation condition and the third excitation condition on the side, the second excitation condition, and the excitation condition corresponding to the middle point of the third excitation condition are the fourth excitation condition. And a difference between the first excitation condition and the fourth excitation condition is compared with a reference value obtained by multiplying the beam focal depth by a constant coefficient, and if the difference is smaller than the reference value, And a fifth step of determining the first excitation condition to be the in-focus point.

Further, according to the present invention, the beam scanning is performed for each excitation condition of the objective lens of the electron microscope, and the obtained secondary electron signal waveform is subjected to line integration to show the value of the line integration as a function of the excitation condition. The first step of forming profile data, the second step of obtaining the half-value width of the profile data, and the half-value width obtained in the second step are within 1/2 of the excitation variable range of the objective lens. And a third step of determining the corresponding case as the in-focus point.

[0014]

According to the first automatic focus determination method of the present invention, the first excitation condition obtained in the second step is set to the third step in correspondence with the peak of the profile data obtained in the first step. Of the profile data obtained in the first step by comparing with the excitation condition corresponding to the midpoint of the second and third excitation conditions obtained in, that is, the fourth excitation condition obtained in the fourth step. When the symmetry between the strong excitation side and the weak excitation side is good, it is determined to be the in-focus state. Therefore, in the fifth step, the difference between the first excitation condition and the fourth excitation condition is It is compared with a reference value obtained by multiplying the beam depth of focus by a constant coefficient.

In the automatic focus determination method according to the second aspect of the present invention, the full width at half maximum of the profile data is obtained in the second step in correspondence with the peak of the profile data obtained in the first step. At this stage, if this half-value width corresponds to ½ or more of the excitation variable range of the objective lens, it is determined that it is out of focus.

[0016]

Embodiments of the present invention will be described below with reference to the drawings. Example 1. FIG. 1 is a sequence diagram for explaining the first embodiment of the automatic focus determination method of the present invention.

First, the scanning speed of the electron beam of the electron microscope is set from the normal 8 kHz to 1 kHz for automatic focusing (step S1).

Next, while sequentially setting the excitation condition of the objective lens (512 conditions) (step S2), scanning is performed four times under each excitation condition (step S3). In this case, 512 samplings are performed per scan.

At the same time, the secondary electron signal obtained for each scan is detected, the data for four scans are added and averaged, and the power supply frequency noise is removed to improve the S / N (step S).
4).

Subsequently, a moving average process is performed on the data subjected to the addition and averaging process to further improve the S / N (step S5).

Next, smoothing and differential processing of 7 points is performed on the data subjected to the moving average processing for each excitation condition, and the sum of the absolute values of the obtained differential coefficient values, that is, the line integral value is obtained (step S6).

The profile composed of the relationship between the integral value obtained as described above and the excitation condition of the objective lens is
As shown in FIG. The maximum value of this profile, that is, the excitation condition of the objective lens corresponding to the peak is obtained, and this excitation condition is set in the objective lens (step S7).

Next, it is judged whether or not the S / N ratio of the integrated value profile data obtained in step S7 is smaller than 3 (step S8).

If it is smaller (Yes), the operation speed is decreased to 0.5 kHz, the number of operations under the excitation condition is increased to 8 (step S9), and the process returns to step S1. Therefore, the numerical values are changed in steps S1 and S3.

If it is larger (No), the next step S
Go to 10. In step S10, it is determined whether or not the excitation condition indicated by the maximum value of the line integral exists in the region of 25% or less of the excitation variable range, or in the region of 75% or more.

If it exists (Yes), the excitation condition indicated by the maximum value is reset to the center of the excitation variable range, and the process returns to step S1.

If it does not exist (No), the process proceeds to the next step S12. In step S12, it is determined whether or not the difference between the excitation condition of the maximum value of the line integral and the center of the excitation condition obtained from the threshold value is equal to or larger than the reference value obtained from the depth of focus.

If it is greater than or equal to the reference value (Yes), the process returns to step S11.

If it is less than the reference value (No), the process proceeds to step S13. In step S13, the maximum value of the previously obtained line integral is determined to be appropriate and set to the objective lens.

Next, the procedure for determining whether or not the focused state is executed in step S12 of FIG. 1 will be described.

For a profile as shown in FIG.
The excitation condition (p) point of the objective lens corresponding to the peak, that is, the focus setting value, and the half value of the peak value (50
%) Of the objective lens excitation condition. In FIG.
The excitation variable range is shown as between A and B. It should be noted that as the half value of the peak value, there is one strong excitation side and one weak excitation side of the objective lens, and both of them are determined as points (a) and (b), respectively.

Next, the excitation condition point (c) of the objective lens corresponding to the midpoint between points (a) and (b) is determined.

Then, the difference df between the point (p) corresponding to the peak and the point (c) corresponding to the middle point is determined as df = | p−c | (1).

The profile ideally shows a symmetrical shape, and in the case of a perfect symmetrical shape, df = 0 (2).

When the expression (2) is satisfied, that is, when the points (c) and (p), which are the midpoints of the points (a) and (b), have substantially one value, they are in focus. Can be said.

On the other hand, when the profile deviates from the symmetrical shape, it can be said that it is out of focus. The reason why the profile is asymmetric is considered to be as follows. a) Noise with a constant cycle (noise due to the power cycle) is mixed in the profile. b) Charge-up occurs due to, for example, the structure of the sample during the execution of the beam scanning for automatic focusing. c) Depending on the case, the influence of the hysteresis of the objective lens is mixed.

The asymmetric state of such a profile is
When the amount exceeds a certain amount, for example, as shown in FIG. 3, variation in dimensional measurement becomes large.

Therefore, when the df value is equal to or larger than a certain value, it is determined that the focus state is out of focus.

If it is determined that the object is out of focus,
Returning to step S11, the focusing sequence is executed again.

Actually, as a method of judging the adequacy of the df value, the value of the depth of focus of the beam is compared with a value obtained by multiplying a certain coefficient, and if it is larger than this, it is regarded as out-of-focus and falls within this range. If so, the focus is determined. The coefficient used in this case varies depending on the pattern structure, magnification, etc., but it is desirable to use a value of 1.0 to 5.0 as a value to be multiplied by the beam depth of focus. Generally, the shallower the depth of focus, the smaller this coefficient is used. A coefficient of 2 μm or less is preferable for a short focus of about 1 μm, and a coefficient of about 4 is preferable for a normal focus of 2 to 3 μm. Also, the pattern structure is
With a resist pattern on Al in a semiconductor wafer,
When the magnification is 30 k times, it is preferable to adopt 3.0 as the coefficient for multiplying the beam depth of focus.

As described above, in the measurement of the resist line pattern having the pattern dimension of the micron level on the semiconductor wafer, the success rate of the automatic focusing is 96.7% (the number of successes is 58 times / execution) by the conventional method. From the number of times 60) to 100% (the number of successes 90 times / the number of execution times 90) in the present invention, it has become possible to dramatically increase. By the way, in this case, the determination of successful focusing is
The variation in dimension measurement after execution of automatic focusing is 1/100 of the pattern design dimension, that is, 0.01 micrometer or less.

As described above, when relying on the operator's visual judgment as the focus determination as in the conventional case, 6
In contrast to 0 trials, an object with improper focus is judged to be in-focus 2 times, but as in the present invention,
After performing the automatic focusing, whether the focusing is successful or not is automatically determined, and when the non-focusing is determined, the automatic focusing is performed again.
% Focus state could be obtained.

In the first embodiment, in the profile data of FIG. 2, the point (a) on the strong excitation side and the point (b) on the weak excitation side corresponding to the peak are 50% of the peak value.
However, this ratio can be appropriately set within a range of about 20% to 80% of the peak value, and it is an empirical factor, the characteristics of the objective lens of the electron microscope, etc. It may be appropriately determined in consideration of various requirements.

Also, as the coefficient for multiplying the beam depth of focus, the range of 1.0 to 5.0 is exemplified in the first embodiment, but this is also an empirical factor, and other coefficients may be used depending on various requirements. It may be used. Embodiment 2. FIG. In the above embodiment, the method for obtaining the peak value and the excitation condition for giving a certain ratio of the peak value from the profile data, and the automatic focus determination method based on these relationships have been described. The half-value width of the profile data is calculated, and if this half-value width is within 1/2 of the excitation variable range of the objective lens of the electron microscope, it is determined to be the in-focus point, and if it is more than that, it is determined to be the non-in-focus point. Also, the same effect can be obtained. In this case, steps similar to step S8 and step S10 shown in FIG. 1 can be provided between the step of obtaining profile data and the step of determining whether or not the focus is on. That is, similarly to step 8, it is determined whether the S / N ratio of the line integral profile data is 3 or less. When it is below, the scanning speed is set to 0.5 as in step S9.
After dropping to kHz, the number of scanning lines is set to 8 and the profile data is taken again. In the above case, the maximum value of the line integral is 25% or less of the excitation variable range, or 75
It is determined whether or not it is in any of the areas of% or more. If it exists in either of the above two areas, step S1
Similar to 1, the maximum excitation condition is reset to the center of the excitable range, and the profile data is collected again. If it does not exist, the focusing point is determined based on the above-described half-width.

Further, in each of the above-described embodiments, the sequence for determining the automatic focusing on the basis of the symmetry of the excitation condition with the peak value of the profile data as the center has been described.
In principle, it is based on the fact that the amount of distortion of profile data is small in the case of focusing, and it goes without saying that any other mathematical method can be applied if the amount of distortion and symmetry of profile data can be determined. Yes.

[0046]

As described above, according to the automatic focus determination method of the present invention, after performing the automatic focusing, the in-focus determination is automatically performed from the obtained profile state, and it is determined that the focus is not in focus. In this case, it is configured to perform automatic focusing again when it is used, so it is possible to dramatically improve the reliability of automatic focusing in electron microscopes, etc., and measure pattern dimensions in semiconductor steps using electron microscopes. There is an effect that the accuracy of can be improved and the efficiency and productivity of the line can be significantly improved.

[Brief description of drawings]

FIG. 1 is a sequence diagram illustrating an embodiment of a focus automatic determination method of the present invention.

FIG. 2 is a profile diagram for explaining the focus automatic determination method of FIG.

FIG. 3 is an explanatory diagram of a result of performing dimension measurement of a pattern formed on a semiconductor chip after performing general automatic focusing.

Claims (11)

[Claims] 1. A profile data showing a value of line integration as a function of the excitation condition is formed by performing beam scanning for each excitation condition of an objective lens of an electron microscope and line-integrating the obtained secondary electron signal waveform. And a second step of obtaining a first excitation condition corresponding to the peak of the profile data, and an excitation condition corresponding to a constant ratio of the peak value of the profile data, A third step of obtaining the second excitation condition and the third excitation condition on the strong excitation side and the weak excitation side, respectively, the second excitation condition, and the excitation corresponding to the middle point of the third excitation condition. A fourth step of obtaining a condition as a fourth excitation condition, and a difference between the first excitation condition and the fourth excitation condition,
Compared with a reference value that is the beam depth of focus multiplied by a constant factor,
If it is smaller than this, a fifth step of determining the first excitation condition as an in-focus point, and a focus automatic determination method characterized by the following. 2. The automatic focus determination method according to claim 1, wherein the first step includes a procedure of setting an excitation condition of the objective lens, beam scanning, and capturing of a secondary electron signal. 3. S / N of the profile data obtained
3. The automatic focus determination method according to claim 1, wherein when the ratio is low, the scanning speed is reduced, the number of scanning lines is increased, and the process returns to the first step. 4. The automatic focus determination method according to claim 1, wherein a ratio within a range of 20% to 80% is used as the constant ratio in the third step. 5. In the fifth step, as the constant coefficient, a smaller value is used as the depth of focus is shallower.
1 to 4 are automatic focus determination methods. 6. The automatic focus determination method according to claim 1, wherein a value of 1.0 to 5.0 is used as the constant coefficient in the fifth step. 7. When the first excitation condition comes near both ends of the excitation variable range, the first excitation condition is newly set at the center of the excitation variable range and the process returns to the first step. An automatic focus determination method according to claim 1. 8. In the fifth step, when the difference is larger than the reference value, it is determined that the focus is not in focus and the previously obtained maximum excitation condition is set at the center of the excitation variable range. Then, the focus automatic determination method according to any one of claims 1 to 7, which returns to the first step. 9. A profile data showing a value of line integration as a function of the excitation condition is formed by performing beam scanning for each excitation condition of the objective lens of the electron microscope and line-integrating the obtained secondary electron signal waveform. The first step, the second step of obtaining the half-value width of the profile data, and the half-value width obtained in the second step being less than or equal to 1/2 of the excitation variable range of the objective lens. A focus automatic determination method comprising: a third step of determining a focus. 10. During the first step and the second step, it is determined whether the S / N ratio of the profile data is lower than a predetermined value. If the S / N ratio is lower than a predetermined value, the scanning speed is reduced and scanning is performed. 10. The automatic focus determination method according to claim 9, further comprising increasing the number of, and returning to the first step. 11. A first excitation condition corresponding to a peak of the profile data is obtained, and when the first excitation condition comes near both ends of the excitation variable range, the first excitation condition is newly set. 11. The automatic focus determination method according to claim 9, wherein the focus is set at the center of the excitation variable range and the process returns to the first step.